Text to video
Describe the scene and choose a 3–30 second duration. For text prompts, select 16:9, 9:16, 1:1 or 21:9.
Create Kling 4.0 AI video from a prompt, guide it with first, middle and last frames, or combine image and video references. This Kling 4.0 generator supports 3–30 second clips, 720p or 1080p output and an aspect ratio suited to your scene.
Describe the scene and choose a 3–30 second duration. For text prompts, select 16:9, 9:16, 1:1 or 21:9.
Upload a first frame, up to eight middle frames and an optional last frame. Refer to the images in your prompt to guide the sequence.
Add up to ten images and five video clips. Use @Image1 or @Video1 in the prompt to identify a specific upload, and choose whether to keep the source video's sound.
This page offers text-to-video, keyframe-guided image-to-video and image/video reference generation. Switch to Kling 3.0 in the model selector if you need its separate controls.
The Kling 4.0 form accepts 3 to 30 seconds. Reference videos are verified after upload; their lengths can affect the billed duration shown before submission.
Yes. Image to video supports a first frame, up to eight middle frames and an optional last frame. The uploaded frames remain visible in the editor.
Type @ in the prompt to open the asset picker, then select the image or video you want. The inserted token identifies that upload.
Edit video and Extend video are marked Coming soon and cannot be submitted yet.